1.
Chapter 2Basic principles of the seismicmethodIn this chapter we introduce the basic notion of seismic waves. In the earth, seismicwaves can propagate as longitudinal (P) or as shear (S) waves. For free space, the one-dimensional wave equation is derived. The wave phenomena occurring at a boundarybetween two layers are discussed, such as Snell’s Law, reﬂection and transmission. Forseismic-exploration purposes, where measurements are taking place at the surface, thediﬀerent arrivals of direct waves, reﬂected waves and refracted/head waves are discussed.2.1 IntroductionThe seismic method makes use of the properties of the velocity of sound. This velocityis diﬀerent for diﬀerent rocks and it is this diﬀerence which is exploited in the seismicmethod. When we create sound at or near the surface of the earth, some energy will bereﬂected back (bounced back). They can be characterized as echoes. From these echoeswe can determine the velocities of the rocks, as well as the depths where the echoes camefrom. In this chapter we will discuss the basic principles behind the behaviour of sound insolid materials. When we use the seismic method, we usually discuss two types of seismicmethods, depending on whether the distance from the sound source to the detector (the”ear”) is large or small with respect to the depth of interest: the ﬁrst is known as refractionseismics, the other as reﬂection seismics. Of course, there is some overlap between thosetwo types and that will be discussed in this chapter. When features really diﬀer, then thatwill be discussed in next chapter for refraction and the chapters on reﬂection seismics.The overlap lies in the physics behind it, so we will deal with these in this chapter. In thefollowing chapters will deal with instrumentation, ﬁeld techniques, corrections (which arenot necessary for refraction data) and interpretation. 18
2.
2.2 Basic physical notions of wavesEverybody knows what waves are when we are talking about waves at sea. Sound inmaterials has the same kind of behaviour as these waves, only they travel much fasterthan the waves we see at sea. Waves can occur in several ways. We will discuss two ofthem, namely the longitudinal and the shear waves. Longitudinal waves behave like wavesin a large spring. When we push from one side of a spring, we will observe a wave goingthrough the spring which characterizes itself by a thickening of the wires running throughthe spring in time (see also ﬁgure (2.1)). A property of this type of wave is that themotion of a piece of the wire is in the same direction as the wave moves. These waves arealso called Push-waves, abbreviated to P-waves, or compressional waves. Another typeof wave is the shear wave. A shear wave can be compared with a chord. When we pusha chord upward from one side, a wave will run along the chord to the other side. Themovement of the chord itself is only up- and downward: characteristic of this wave is thata piece of the chord is moving perpendicular to the direction of that of the wave (seealso ﬁgure (2.1)). These types of waves are referred to as S-waves, also called dilatationalwaves. Characteristic of this wave is that a piece of the chord is pulling its ”neighbour”upward, and this can only occur when the material can support shear strain. In ﬂuids,one can imagine that a ”neighbour” cannot be pulled upward simply because it is a ﬂuid.Therefore, in ﬂuids only P-waves exist, while in a solid both P- and S-waves exist. Waves are physical phenomena and thus have a relation to basic physical laws. Thetwo laws which are applicable are the conservation of mass and Newton’s second law.These two have been used in appendix A to derive the two equations governing the wavemotion due to a P-wave. There are some simplifying assumptions in the derivation, oneof them being that we consider a 1-dimensional wave. When we denote p as the pressureand vx as the particle velocity, the conservation of mass leads to: 1 ∂p ∂vx =− (2.1) K ∂t ∂xin which K is called the bulk modulus. The other relation follows from application ofNewton’s law: ∂p ∂vx − =ρ (2.2) ∂x ∂twhere ρ denotes the mass density. This equation is called the equation of motion. Thecombination of these two equations leads, for constant density ρ, to the equation whichdescribes the behaviour of waves, namely the wave equation: ∂2p 1 ∂2p − 2 2 =0 (2.3) ∂x2 c ∂tin which c can be seen as the velocity of sound, for which we have: c = K/ρ. The 19
3.
Figure 2.1: Particle motion of P and S wavessolution to this equation is: p(x, t) = s(t ± x/c) (2.4)where s(t) is some function. Note the dependency on space and time via (t ± x/c), whichdenotes that it is a travelling wave. The sign in the argument is depending on whichdirection the wave travelling in. Often, seismic responses are analyzed in terms of frequencies, i.e., the Fourier spectraas given in Chapter 1. The deﬁnition we use here is: +∞ G(ω) = g(t) exp(−iωt) dt (2.5) −∞which is the same as equation (1.1) from Chapter 1 but we used the radial frequency ωinstead: ω = 2πf. Using this convention, it is easy to show that a diﬀerentiation withrespect to time is equivalent to multiplication with iω in the Fourier domain. When wetransform the solution of the wave equation (equation(2.4)) to the Fourier domain, weobtain: P (x, ω) = S(ω) exp(±iωx/c). (2.6) 20
4.
Note here that the time delay x/c (in the time domain) becomes a linear phase in theFourier domain, i.e., (ωx/c). In the above, we gave an expression for the pressure, but one can also derive theequivalent expression for the particle velocity v x . To that purpose, we can use the equationof motion as expressed in equation (2.2), but then in its Fourier-transformed version, whichis: 1 ∂P (x, ω) Vx (x, ω) = − . (2.7) iωρ ∂xWhen we subsitute the solution for the pressure from above (equation (2.6)), we get forthe negative sign: 1 −iω Vx (x, ω) = − S(ω) exp(−iωx/c) iωρ c 1 = S(ω) exp(−iωx/c). (2.8) ρcNotice that the particle velocity is a scaled version of the pressure: P (x, ω) Vx (x, ω) = . (2.9) ρcThe scaling factor is (ρc), being called the seismic impedance. In the previous analysis, we considered one-dimensional waves. Normally in the realworld, we deal with three dimensions, so a wave will spread in three directions. In ahomogeneous medium (so the properties of the material are everywhere constant and thesame) the wave will spread out like a sphere. The outer shell of this sphere is called thewave front. Another way of describing this wave front is in terms of the normal to thewavefront: the ray. We are used to rays in optics and we can use the same notion inthe seismic method. When we were explaining the behaviour of P- and S-waves, we arealready using the term ”neighbour”. This was an important feature otherwise the wavewould not move forward. A fundamental notion included in this, is Huygens’ principle.When a wave front arrives at a certain point, that point will behave also as a source forthe wave, and so will all its neighbours. The new wavefront is then the envelope of all thewaves which were generated by these points. This is illustrated in ﬁgure (2.2). Again, theray can then be deﬁned as the normal to that envelope which is also given in the ﬁgure. So far, we only discussed the way in which the wave moves forward, but there is alsoanother property of the wave we haven’t discussed yet, namely the amplitude : how doesthe amplitude behave as the wave moves forward? We have already mentioned sphericalspreading when the material is everywhere the same. The total energy will be spreadedout over the area over the sphere. This type of energy loss is called spherical divergence.It simply means that if we put our ”ear” at a larger distance, the sound will be less loud. 21
5.
secondary sources t at t = t0 wav efron c∆ t = t0 + ∆t fr on t at t wave Figure 2.2: Using Huygens’ principle to locate new wavefronts. Material velocity Material velocity (m/s) (m/s) Air 330 Sandstone 2000-4500 Water 1500 Shales 3900-5500 Soil 20-300 Limestone 3400-7000 Sands 600-1850 Granite 4800-6000 Clays 1100-2500 Ultra-maﬁc rocks 7500-8500 Table 2.1: Seismic wave velocities for common materials and rocks.There is also another type of energy loss, and that is due to losses within the material,which mainly consists of internal friction losses. This means that the amplitude of a wavewill be extra damped because of this property. S-waves usually show higher friction lossesthan P-waves. Finally, we give a table of common rocks and their seismic wave velocitiesin table (2.1).2.3 The interface : Snell’s law, refraction and reﬂectionSo far, we discussed waves in a material which had everywhere the same constant wavevelocity. When we include a boundary between two diﬀerent materials, some energy isbounced back, or reﬂected, and some energy is going through to the other medium. Itis nice to perform Huygens’ principle graphically on such a conﬁguration to see how thewavefront moves forward (propagates), especially into the second medium. From this 22
6.
A" 0 e im tt t ta e im n ro tt ef ta av on w θ1 r ef A’ av w velocity c1 A B’ velocity c2 θ2 B Figure 2.3: Snell’s lawpicture, we could also derive the ray concept. In this discussion, we will only considerthe notion of rays. A basic notion in the ray concept, is Snell’s law. Snell’s law is afundamental relation in the seismic method. It tells us the relation between angle ofincidence of a wave and velocity in two adjacent layers (see Figure (2.3)). AA A is part of a plane wave incident at angle θ 1 to a plane interface between medium1 of velocity c1 , and medium 2 of velocity c2 . The velocities c1 and c2 are constant. In atime t the wave front moves to the position AB and are normals to the wave front. So thetime t is given by AB AB t= = (2.10) c1 c2Considering the two triangles and this may be written as: AB sin θ1 AB sin θ2 t= = (2.11) c1 c2 23
7.
Hence, sin θ1 sin θ2 = (2.12) c1 c2which is Snell’s law for transmission. So far, we have taken general velocities c 1 and c2 .However, in a solid, two velocities exist, namely P- and S-wave velocities. Generally, whena P-wave is incident on a boundary, it can transmit as a P-wave into the second medium,but also as a S-wave. So in the case of the latter, Snell’s law reads: sin θP sin θS = (2.13) cP cSwhere cP is the P-wave velocity, and cS the S-wave velocity. The same holds for reﬂection:a P-wave incident on the boundary generates a reﬂected P-wave and a reﬂected S-wave.Finally, the same holds for an incident S-wave: it generates a reﬂected P-wave, a reﬂectedS-wave, a transmitted P-wave and a transmitted S-wave. A special case of Snell’s law is of interest in refraction prospecting. If the ray isrefracted along the interface (that is, if θ 2 = 90 deg), we have sin θc 1 = (2.14) c1 c2where θc is known as the critical angle. So far, we have looked at basic notions of refraction and reﬂection at an interface.When we measure in the ﬁeld, and there would be one boundary below it, we couldobserve several arrivals: a direct ray, a reﬂected ray and a refracted ray. We will derivethe arrival time of each ray as depicted in ﬁgure 2.4. The direct ray is very simple: it is the horizontal distance divided by the velocity ofthe wave, i.e.,: x t= (2.15) c1When we look at the reﬂected ray, we have that the angle of incidence is the same as theangle of reﬂection. This also follows from Snell’s law: when the velocities are the same,the angles must also be the same. When we use Pythagoras’ theorem, we obtain for thetraveltime: 1/2 4z 2 + x2 t= (2.16) c1When we square this equation, we see that it is the equation of a hyperbola: 2 2 2z x t2 = + (2.17) c1 c1 24
8.
distance x A D direct thickness z re fle ct io n velocity c1 B refraction C velocity c2 Figure 2.4: The direct, reﬂected and refracted ray. When we look at the refracted ray the derivation is a bit more complicated. Take eachray element, so take the paths AB, BC and CD separately. Then, for the ﬁrst element,as shown in ﬁgure (2.5), we obtain the traveltime: ∆s1 + ∆s2 ∆x1 sin θc z cos θc ∆t1 = = + (2.18) c1 c1 c1where θc is the critical angle. We can do this also for the paths BC and CD, and we obtain the total time as: t = ∆t1 + ∆t2 + ∆t3 (2.19) ∆x1 sin θc z cos θc ∆x2 ∆x3 sin θc z cos θc = + + + + (2.20) c1 c1 c2 c1 c1where ∆x2 = BC and ∆x3 is the horizontal distance between C and D. When we usenow Snell’s law, i.e., sin θc /c1 = 1/c2 , in the terms with ∆x1 and ∆x3 , then we can addall the terms with 1/c2 , using x = ∆x1 + ∆x2 + ∆x3 to obtain: x 2z cos θc t= + (2.21) c2 c1We recognize this equation as the equation of a straight line when t is considered as afunction of distance x, the line along which we do our measurements. We will use this 25
9.
A ∆s 1 θc z ∆s 2 ∆x 1 B refracted ray Figure 2.5: An element of the ray with critical incidenceequation later in the next chapter. We have now derived the equations for the three rays,and we can plot the times as a function of x. This is done in ﬁgure (2.6). This picture is an important one. When we measure data in the ﬁeld the characteristicsin this plot can most of the time be observed. Now we have generated this ﬁgure, we can specify better when we are performinga reﬂection survey, or a refraction survey. In refraction seismics we are interested inthe refractions and only in the travel times. This means that we can only observe thetraveltimes well if it is not masked by the reﬂections or the direct ray, which meansthat we must measure at a relatively large distance with respect to the depth of interest.This is diﬀerent with reﬂection seismics. There, the reﬂections will always be masked byrefractions or the direct ray, but there are ways to enhance the reﬂections. What we areinterested in, is the arrival at relatively small oﬀsets, thus distances of the sound sourceto the detector which are small with respect to the depth we are interested in. Before discussing any more diﬀerences between the refraction method and the reﬂectionmethod, we would like to discuss amplitude eﬀects at the boundary. Let us ﬁrst introducethe acoustic impedance, which is the product of the density ρ with the wave velocity c,i.e., ρc. When a ray encounters a boundary, some energy will be reﬂected back, and somewill be transmitted to the next layer. The amount of energy reﬂected back is characterizedby the reﬂection coeﬃcient R: ρ2 c2 − ρ 1 c1 R= (2.22) ρ2 c2 + ρ 1 c1 26
10.
traveltime n on io ti rac ct ref fle re t ec dir 0 distance x Figure 2.6: Time-distance (t, x) curve for direct, reﬂected and refracted ray.That this is the case, will be derived from basic physical principles in the chapter onprocessing (Chapter 5). Obviously, the larger the impedance contrast between two layers,the higher the amplitude of the reﬂected wave will be. Notice that it is the impedancecontrast which determines whether energy is reﬂected back or not; it may happen thatthe velocities and densities are diﬀerent between two layers, but that the impedance is(nearly) the same. In that case we will see no reﬂection. We can now state anotherdiﬀerence between refraction and reﬂection seismics. With refraction seismics we are onlyinterested in traveltimes of the waves, so this means that we are interested in contrasts invelocities. This is diﬀerent in reﬂection seismics. Then we are interested in the amplitudeof the waves, and we will only measure an amplitude if there is a contrast in acousticimpedance in the subsurface. Generally speaking the ﬁeld equipment for refraction and reﬂection surveys have thesame functionality: we need a source, detectors and recording equipment. Since reﬂectionseismics gives us a picture of the subsurface, it is much used by the oil industry andtherefore, they put high demands on the quality of the equipment. As said before, adiﬀerence between the two methods is that in reﬂection seismics we are interested inamplitudes as well, so this means that high-precision instruments are necessary to pick 27
11.
REFRACTION SEISMICS REFLECTION SEISMICS Based on contrasts in : Based on contrasts in : seismic wave speed (c) seismic wave impedances (ρc) Material property determined : Material properties determined: wave speed only wave speed and wave impedance Only traveltimes used Traveltimes and amplitudes used No need to record amplitudes completely : Must record amplitudes correctly : relatively cheap instruments relatively expensive instruments Source-receiver distances large compared to Source-receiver distances small compared to investigation depth investigation depth Table 2.2: Important diﬀerences between refraction and reﬂection seismics.those up accurately. Source, detectors and recording equipment will be discussed in thechapter on seismic instrumentation (chapter 3). Finally, we tabulate the most important diﬀerences between reﬂection and refractionseismic in table 2.2. 28
Clipping is a handy way to collect and organize the most important slides from a presentation. You can keep your great finds in clipboards organized around topics.
Be the first to comment